{"total":74,"rows":[{"id":"2406.11717","title":"Refusal in Language Models Is Mediated by a Single Direction","created":"2024-06-17","cited_by_count":1032,"area":"safety"},{"id":"2404.01318","title":"JailbreakBench: An Open Robustness Benchmark for Jailbreaking Large Language Models","created":"2024-03-28","cited_by_count":536,"area":"safety"},{"id":"2312.00273","title":"Mark My Words: Analyzing and Evaluating Language Model Watermarks","created":"2023-12-01","cited_by_count":71,"area":"safety"},{"id":"2402.09179","title":"Instruction Backdoor Attacks Against Customized LLMs","created":"2024-02-14","cited_by_count":62,"area":"safety"},{"id":"2502.19820","title":"Foot-In-The-Door: A Multi-turn Jailbreak for LLMs","created":"2025-02-27","cited_by_count":54,"area":"safety"},{"id":"2501.01765","title":"SaLoRA: Safety-Alignment Preserved Low-Rank Adaptation","created":"2025-01-03","cited_by_count":42,"area":"safety"},{"id":"2502.01042","title":"SafeSwitch: Steering Unsafe LLM Behavior via Internal Activation Signals","created":"2025-02-03","cited_by_count":29,"area":"safety"},{"id":"2406.12257","title":"CleanGen: Mitigating Backdoor Attacks for Generation Tasks in Large Language Models","created":"2024-06-18","cited_by_count":24,"area":"safety"},{"id":"2502.12176","title":"Ten Challenging Problems in Federated Foundation Models","created":"2025-02-14","cited_by_count":20,"area":"safety"},{"id":"2410.02440","title":"Optimizing Adaptive Attacks against Watermarks for Language Models","created":"2024-10-03","cited_by_count":16,"area":"safety"},{"id":"2502.02787","title":"SimMark: A Robust Sentence-Level Similarity-Based Watermarking Algorithm for Large Language Models","created":"2025-02-05","cited_by_count":15,"area":"safety"},{"id":"2502.04204","title":"Short-length Adversarial Training Helps LLMs Defend Long-length Jailbreak Attacks: Theoretical and Empirical Evidence","created":"2025-02-06","cited_by_count":14,"area":"safety"},{"id":"2503.08195","title":"Dialogue Injection Attack: Jailbreaking LLMs through Context Manipulation","created":"2025-03-11","cited_by_count":14,"area":"safety"},{"id":"2506.05735","title":"Do LLMs Really Forget? Evaluating Unlearning with Knowledge Correlation and Confidence Awareness","created":"2025-06-06","cited_by_count":14,"area":"safety"},{"id":"2504.15047","title":"RainbowPlus: Enhancing Adversarial Prompt Generation via Evolutionary Quality-Diversity Search","created":"2025-04-21","cited_by_count":11,"area":"safety"},{"id":"2508.01365","title":"ConfGuard: A Simple and Effective Backdoor Detection for Large Language Models","created":"2025-08-02","cited_by_count":9,"area":"safety"},{"id":"2508.09603","title":"The Surprising Effectiveness of Membership Inference with Simple N-Gram Coverage","created":"2025-08-13","cited_by_count":7,"area":"safety"},{"id":"2601.02751","title":"Window-based Membership Inference Attacks Against Fine-tuned Large Language Models","created":"2026-01-06","cited_by_count":7,"area":"safety"},{"id":"2506.12913","title":"Jailbreak Transferability Emerges from Shared Representations","created":"2025-06-15","cited_by_count":6,"area":"safety"},{"id":"2602.15143","title":"Protecting Language Models Against Unauthorized Distillation through Trace Rewriting","created":"2026-02-16","cited_by_count":6,"area":"safety"},{"id":"2603.01291","title":"JailNewsBench: Multi-Lingual and Regional Benchmark for Fake News Generation under Jailbreak Attacks","created":"2026-03-01","cited_by_count":6,"area":"safety"},{"id":"2405.16783","title":"TrojFM: Resource-efficient Backdoor Attacks against Very Large Foundation Models","created":"2024-05-27","cited_by_count":5,"area":"safety"},{"id":"2411.19335","title":"PEFT-as-an-Attack! Jailbreaking Language Models during Federated Parameter-Efficient Fine-Tuning","created":"2024-11-28","cited_by_count":5,"area":"safety"},{"id":"2605.12456","title":"TextSeal: A Localized LLM Watermark for Provenance & Distillation Protection","created":"2026-05-12","cited_by_count":5,"area":"safety"},{"id":"2503.24191","title":"When Grammar Guides the Attack: Uncovering Control-Plane Vulnerabilities in LLMs with Structured Output","created":"2025-03-31","cited_by_count":4,"area":"safety"},{"id":"2603.06588","title":"vLLM Hook v0: A Plug-in for Programming Model Internals on vLLM","created":"2026-02-02","cited_by_count":4,"area":"safety"},{"id":"2604.08169","title":"Activation Steering for Aligned Open-ended Generation without Sacrificing Coherence","created":"2026-04-09","cited_by_count":4,"area":"safety"},{"id":"2403.06131","title":"PPFedIT: Towards Privacy-Preserving Federated Instruction Tuning with Few-shot Local Examples","created":"2024-03-10","cited_by_count":3,"area":"safety"},{"id":"2409.19173","title":"HM3: Heterogeneous Multi-Class Model Merging","created":"2024-09-27","cited_by_count":3,"area":"safety"},{"id":"2510.13351","title":"Protect: Towards Robust Guardrailing Stack for Trustworthy Enterprise LLM Systems","created":"2025-10-15","cited_by_count":3,"area":"safety"},{"id":"2512.16292","title":"In-Context Probing for Membership Inference in Fine-Tuned Language Models","created":"2025-12-18","cited_by_count":3,"area":"safety"},{"id":"2601.03699","title":"RedBench: A Universal Dataset for Comprehensive Red Teaming of Large Language Models","created":"2026-01-07","cited_by_count":3,"area":"safety"},{"id":"2602.04581","title":"Trust The Typical","created":"2026-02-04","cited_by_count":3,"area":"safety"},{"id":"2604.09544","title":"Large Language Models Generate Harmful Responses Using a Distinct Mechanism, Shared Across Harm Types","created":"2026-04-10","cited_by_count":3,"area":"safety"},{"id":"2605.27825","title":"MRMMIA: Membership Inference Attacks on Memory in Chat Agents","created":"2026-05-27","cited_by_count":2,"area":"safety"},{"id":"2606.19535","title":"FloatDoor: Platform-Triggered Backdoors in LLMs","created":"2026-06-17","cited_by_count":2,"area":"safety"},{"id":"2506.13494","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","created":"2025-06-16","cited_by_count":1,"area":"safety"},{"id":"2507.01752","title":"Tuning without Peeking: Provable Generalization Bounds and Robust LLM Post-Training","created":"2025-07-02","cited_by_count":1,"area":"safety"},{"id":"2603.10068","title":"ADVERSA: Measuring Multi-Turn Guardrail Degradation and Judge Reliability in Large Language Models","created":"2026-03-10","cited_by_count":1,"area":"safety"},{"id":"2604.27861","title":"TwinGate: Stateful Defense against Decompositional Jailbreaks in Untraceable Traffic via Asymmetric Contrastive Learning","created":"2026-04-30","cited_by_count":1,"area":"safety"},{"id":"2605.11653","title":"Every Bit, Everywhere, All at Once: A Binomial Multibit LLM Watermark","created":"2026-05-12","cited_by_count":1,"area":"safety"},{"id":"2605.17075","title":"A Red Teaming Framework for Evaluating Robustness of AI-enabled Security Orchestration, Automation, and Response Systems","created":"2026-05-16","cited_by_count":1,"area":"safety"},{"id":"2605.29434","title":"AliMark: Enhancing Robustness of Sentence-Level Watermarking Against Text Paraphrasing","created":"2026-05-28","cited_by_count":1,"area":"safety"},{"id":"2606.04027","title":"MaskForge: Structure-Aware Adaptive Attacks for Jailbreaking Diffusion Large Language Models","created":"2026-06-01","cited_by_count":1,"area":"safety"},{"id":"2606.09890","title":"PreAct-Bench: Benchmarking Predictive Monitoring in LLMs","created":"2026-06-03","cited_by_count":1,"area":"safety"},{"id":"2606.11425","title":"JailbreakOPT: Tool-Assisted Iterative Jailbreak Prompt Optimization","created":"2026-06-09","cited_by_count":1,"area":"safety"},{"id":"2606.20668","title":"BELLS-O: Evaluating the Operational Trade-offs of LLM Supervision Systems","created":"2026-06-12","cited_by_count":1,"area":"safety"},{"id":"2606.27210","title":"Paved with True Intents: Intent-Aware Training Improves LLM Safety Classification Across Training Regimes","created":"2026-06-25","cited_by_count":1,"area":"safety"},{"id":"2608.27666","title":"Semantic Watermarking with Order-Robust Detection over Sub-sentence Units","created":"2026-08-27","cited_by_count":1,"area":"safety"},{"id":"2609.21793","title":"CASCADE Against Jailbreaks: Combination Across Stages with Controlled Attack-Defense Evaluation","created":"2026-09-18","cited_by_count":1,"area":"safety"},{"id":"2510.23074","title":"Fast-MIA: Efficient and Scalable Membership Inference for LLMs","created":"2025-10-27","cited_by_count":0,"area":"safety"},{"id":"2511.05532","title":"Beyond One-Size-Fits-All: Personalized Harmful Content Detection with In-Context Learning","created":"2025-10-29","cited_by_count":0,"area":"safety"},{"id":"2604.23067","title":"Training a General Purpose Automated Red Teaming Model","created":"2026-04-24","cited_by_count":0,"area":"safety"},{"id":"2605.01350","title":"LLM Output Detectability and Task Performance Can be Jointly Optimized","created":"2026-05-02","cited_by_count":0,"area":"safety"},{"id":"2605.03226","title":"Self-Mined Hardness for Safety Fine-Tuning","created":"2026-05-04","cited_by_count":0,"area":"safety"},{"id":"2605.07032","title":"A Systematic Investigation of RL-Jailbreaking in LLMs","created":"2026-05-07","cited_by_count":0,"area":"safety"},{"id":"2605.22880","title":"How Far Will They Go? Red-Teaming Online Influence with Large Language Models","created":"2026-05-20","cited_by_count":0,"area":"safety"},{"id":"2605.22984","title":"Test-Time Training Undermines Safety Guardrails","created":"2026-05-21","cited_by_count":0,"area":"safety"},{"id":"2605.29659","title":"Opir: Efficient Multi-Task Safety Classification for Toxicity, Jailbreaks, Hate Speech, and Harmful Content","created":"2026-05-28","cited_by_count":0,"area":"safety"},{"id":"2606.29646","title":"Fuzzing Large Language Models to Elicit Hidden Behaviours","created":"2026-06-28","cited_by_count":0,"area":"safety"},{"id":"2607.13088","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","created":"2026-07-13","cited_by_count":0,"area":"safety"},{"id":"2607.22595","title":"xMIx: High-Performance Serving-Time Platform for Mechanistic Interpretability Apps","created":"2026-06-11","cited_by_count":0,"area":"safety"},{"id":"2608.02681","title":"Safety in Batches? Understanding and Mitigating Safety Failures in Batch Prompting","created":"2026-08-03","cited_by_count":0,"area":"safety"},{"id":"2608.05659","title":"Breaking Customized LLMs for Coding: Automated Red Teaming for Instruction Backdoor Attacks","created":"2026-08-06","cited_by_count":0,"area":"safety"},{"id":"2608.30703","title":"SingProbe Technical Report","created":"2026-08-31","cited_by_count":0,"area":"safety"},{"id":"2609.03844","title":"Flip, Don't Shuffle: Watermarking LLMs at the Speed of Inference","created":"2026-09-03","cited_by_count":0,"area":"safety"},{"id":"2609.05794","title":"Bait-and-Recover: Poisoning Internal Refusal Signals to Defend LLMs against White-Box Editing Jailbreaks","created":"2026-09-05","cited_by_count":0,"area":"safety"},{"id":"2609.06966","title":"MOLE: Detecting Insider Threats in AI Agents","created":"2026-09-07","cited_by_count":0,"area":"safety"},{"id":"2610.01768","title":"The Innocent Courier: Covert Exfiltration Through Legitimate LLM Web Fetching","created":"2026-10-01","cited_by_count":0,"area":"safety"},{"id":"2610.06064","title":"TrustMI: Causally controlling how assistants trust their users","created":"2026-10-05","cited_by_count":0,"area":"safety"},{"id":"2610.18359","title":"Pin It on Anyone: Certified Targeted Displacement of Multi-Bit LLM Watermarks Without the Key","created":"2026-10-13","cited_by_count":0,"area":"safety"},{"id":"2610.18588","title":"Decomposing and Predicting Annotation Disagreement in Harmful Content Across Languages","created":"2026-10-13","cited_by_count":0,"area":"safety"},{"id":"2610.19133","title":"TRACER: Trace-Regularized Training Against Indirect Prompt Injection","created":"2026-10-13","cited_by_count":0,"area":"safety"},{"id":"2610.19306","title":"Did the LLM Answer the Question? A Semantic Reconstruction Game for Jailbreak Evaluation","created":"2026-10-13","cited_by_count":0,"area":"safety"}],"next_cursor":null}